Deference and the end of tax practice

Gans, Mitchell M

Editors' Synopsis: The author examines the various standards of deference the courts apply in the case of regulations, revenue rulings and the government's litigation arguments. The article first focuses on a generation skipping transfer tax issue to illustrate how the government recently relied on expanding notions of deference to resolve a conflict in the circuit courts-in effect declaring victory by regulation. It concludes with an argument for legislation that would reduce the level of deference the government currently enjoys and that would preclude it from disavowing taxpayer-friendly rulings or regulations.

INTRODUCTION

Early in the history of the income tax, the Supreme Court indicated that when the meaning of tax legislation is uncertain, taxpayers ought to receive the benefit of the doubt.1 Fifteen years later the Court retreated from this position, concluding that the function of the courts is to determine the meaning of ambiguous legislation and that this function should be performed without giving either party a presumptive advantage The Court has now come full circle, except that instead of granting a preference to taxpayers in the case of uncertainty,3 it grants a preference to the government under the guise of deference-whether to a regulation,4 a revenue ruling, or perhaps even an argument made by the government during litigation.5

Recent developments, particularly in the transfer-tax area, suggest this transformation has not been lost on the government. Part I of the Article sets the stage for a discussion of deference in the tax context by focusing on a recently litigated generation-skipping-tax issue that resulted in a conflict in the circuit courts.6 Rather than taking the traditional route of continuing to litigate in other circuits or seeking Supreme Court review, the government resolved the issue in its favor by promulgating an amendment to the regulations.7 The government issued the amendment approximately fourteen years after the enactment of the underlying statute, after the government had secured a victory in one circuit by persuading the court to uphold the original regulations partly because Congress had ratified them, and after the issue already had been resolved against the government in another circuit. In effect, the government declared victory by its own regulation.

Under the traditional view, the adoption of regulations only after the outbreak of litigation or the failure to adopt regulations contemporaneously with the enactment of the statute to which they relate would have cut against deference. The deference formulation now used by the courts makes these considerations irrelevant. Under the Supreme Court's Chevron doctrine, the agencies are permitted to resolve questions of statutory ambiguity, and the courts are required to defer to any reasonable resolution on the theory that, unlike the courts, the agencies are politically accountable.8

In a 1996 nontax case, the Court invoked Chevron and deferred to an agency interpretation issued approximately one hundred years after the enactment of the statute and after litigation in the lower courts had produced conflicting decisions about the meaning of the statute. And, in a 1997 transfer-tax case, a four-Justice plurality opinion suggested that the court's conclusion could be overturned by new regulations, even though Congress had enacted the statute some fifty years earlier (and the government issued such regulations in 2000).9 The Court's deference formulation even permits government agencies to modify their interpretations over time if the modification is justified, with various Justices intimating that a change in administration might provide sufficient justification. Under this evolving approach, statutes no longer have a fixed meaning. Instead, statutory content can change as the administration, or the philosophy of the agencies, changes. Part II examines and critiques the new deference formulation.

Part III of the Article revisits the generation-skipping-tax issue examined in Part I. It considers whether the regulations issued by the government in the aftermath of the inter-circuit conflict can be sustained, given the government's successful argument that Congress ratified the original regulations under the so-called reenactment doctrine. Even under the expansive conception of deference currently in vogue, it is possible that regulations cannot be altered once Congress has approved them. Part III also examines the relationship between deference and the reenactment doctrine, raising the question of whether newly issued generation-skipping tax regulations are valid.

Although the Supreme Court has not spoken as definitively about the level of deference that revenue rulings are entitled to receive, recent developments suggest that the Chevron framework does not apply in this context. Instead, it appears, a more flexible type of deference is to be applied depending upon a variety of factors that would be irrelevant under Chevron. So, for example, whereas a ruling designed to influence the outcome of pending litigation would probably receive little, or no, deference, a regulation issued in the same circumstances would, in all likelihood, be binding on the courts (assuming it otherwise satisfied Chevron's criteria). At the same time, the government enjoys a different kind of deference that is more analogous to Chevron's framework where a regulation is ambiguous. In this context, a ruling, or perhaps even a government argument made during litigation, that resolves the ambiguity in a reasonable manner may be entitled to Chevron-type binding deference. Part IV of this Article examines the evolving standards of deference applicable to rulings, and Part V considers the issue in the context of interpretations offered by the government with regard to ambiguous regulations.

While, as suggested, the deference standard will differ depending upon the kind of interpretation invoked by the government, the transformation in deference has a theme that cuts across all of the standards. In each case, the law appears to be developing in the direction of granting the government more deference. This is not surprising given Chevron's basic insight: the agencies' political accountability makes them a more appropriate repository for interpretive responsibility than the courts. In the tax context, the transformation portends the end of practice as traditionally understood. If the government can declare victory by regulation, the balance between the government and the taxpayer has been radically changed.

Part VI of the Article argues for a return to the traditional view, under which the government would receive a reduced level of deference. More specifically, it recommends that Chevron should not apply in tax cases. This is not to suggest that the government should receive no deference in the case of a regulation. Rather, the courts should be permitted to take into account various factors when determining the validity of a regulation, such as whether the government issued a regulation contemporaneously with the enactment of the statute or issued it in the heat of litigation. On the other hand, revenue rulings and government-proffered interpretations of ambiguous regulations should receive no deference.

Part VI also considers taxpayer-friendly regulations and rulings. The courts have indicated that, as a matter of constitutional law, a government interpretation deemed inconsistent with the statute must be disregarded when it favors the taxpayer. Because of the unfairness and inefficiency of this approach, Part VI suggests legislation that would circumvent this constitutional limitation by denying funds to the government for the enforcement of any court decision invalidating a taxpayer-friendly interpretation.

PART I. GENERATION-SKIPPING TRANSFER TAX: HISTORY BACKGROUND, AND NEW REGULATIONS

A. History and Background

In 1976, Congress created the generation-skipping transfer tax ("GST").10 Sensitive to concerns about wealth transfers made in reliance on preexisting law, Congress provided grandfather protection for previously created irrevocable trusts (i.e., for trusts that were irrevocable on June 11, 1976).11 In 1986, Congress repealed the GST, making the repeal retroactive to the date of the original enactment.12 Congress simultaneously created a new GST, similar in some respects to its predecessor but critically different in other respects. The new GST also grandfathered preenactment irrevocable trusts (in this case, trusts that were irrevocable on September 25, 1985).13 The grandfather provision in both versions of the GST tax was subject to an exception: the GST would be applicable to a preexisting trust to the extent of any post-enactment addition to the trust. 14

In 1980, the Department of the Treasury ("Treasury") finalized the effective-date regulations under the 1976 legislation.15 In the preamble to the regulations, the Treasury discussed comments it had received regarding preexisting trusts that contain powers of appointment. 16 The commentators had argued that transfers occurring as a result of the exercise or lapse of a power should be grandfathered because the exception from the grandfather provision for post-enactment additions to a trust contemplated additions from outside sources.17 The Treasury rejected this argument. Citing to the Conference Report and pointing to the structure of the statute, the Treasury articulated Congress' rationale for grandfather protection and the exception for additions: to exclude from tax any transfer that could not be undone at the time of enactment (in other words, not to violate reliance interests).18 The regulations invoked traditional estate and gift tax principles to implement this concept. Thus, under the regulations, an addition to the trust would not be viewed as occurring unless a taxable transfer (under either the estate or gift tax) had been made.19

Applying this concept in the context of powers of appointment, the regulations contain a constructive-addition provision.20 Accordingly, transfers occurring as a result of the post-enactment lapse or exercise of a special power of appointment-not being a taxable event under traditional estate and gift tax principles-would not be subject to the GST." However, referencing the 1976 legislative history, the Treasury observed that Congress explicitly indicated its intention to make the post-enactment exercise of a special power under a pre-enactment trust subject to the GST if exercised in a manner that violates a federalized version of the rule against perpetuities. Based on this observation, the Treasury qualified the constructive-addition provision, providing that transfers made under an exercise of a special power of appointment in violation of a federalized perpetuities rule would be subject to the GST.

When Congress adopted the new version of the generation skipping tax in 1986, it explicitly focused on the regulation containing the constructive-- addition provision.24 In both the House and Senate, the conclusion was reached that the regulation accurately captured Congress' intent underlying the 1976 legislation.25 It was agreed, moreover, that the provision should be ratified and remain viable under the new version of the tax.26 Thus, Congress, in 1986, clearly embraced the notion that an addition would render grandfather protection unavailable only where a taxable transfer under the estate or gift tax provisions occurred (subject to the same qualification contained in the 1980 regulations regarding the exercise of a special power of appointment beyond a federalized perpetuities period).27

B. Conflict in the Circuit Courts

In Peterson,28 the husband had created an irrevocable trust prior to the enactment of the 1986 version of the GST.29 Under the trust, the wife had a testamentary general power of appointment." If she failed to exercise the power, the corpus would remain in the trust for the benefit of the grandchildren, except for the portion necessary to pay the estate tax attributable to the inclusion of the corpus in the wife's estate.31 In 1987, the wife died without having exercised the power. The Internal Revenue Service ("Service") took the position that the lapse of the wife's power of appointment upon her death constituted a constructive addition to the trust, thereby making grandfather protection unavailable.33

The Service argued that one of the examples in the then temporary regulations illustrating the constructive-addition provision was directly on point. In the example, the post-enactment lapse of a general power of appointment created under a pre-enactment trust negated grandfather protection.35 The parties agreed that if the constructive-addition regulation were valid, the lapse of the power subjected the transfer to the GST.36 Arguing against the validity of the regulation, the taxpayer made the same argument that the Treasury had rejected in the 1980 preamble to the regulations under the 1976 legislation: the addition concept contained in the statute contemplated transactions having the effect of increasing the amount of trust corpus. The Service argued that the regulation was a reasonable construction of the statute and that it could not be invalidated given that "the regulation received an overall endorsement" by Congress in 1986.38

After reviewing the history of the regulation, the Peterson court agreed with the Service.39 The court first considered the 1980 preamble, in which the Treasury explained that a constructive addition would be deemed to occur only if a transfer taxable for estate or gift tax purposes were made.'40 The court concluded that the Treasury's decision in 1980 to ground the statutory concept of an addition in traditional transfer tax theory was appropriate." Congress had used the term "added"' against the backdrop of the provisions in the Code and regulations that define a taxable transfer, and, therefore, must have intended to give the term content based on these provisions. Because the lapse of a general power of appointment is a taxable event under either the estate43 or gifts tax, the court could have ruled for the Service on this basis and ended its analysis. The Court, however, chose not to do so. Instead, the court concluded that, Congress did not disclose any intent to overrule the regulation 45 Indeed, according to the Court, the legislative history revealed what the Service had argued-that Congress intended to ratify the regulation and to adopt it as the governing construction of the effective-date provision under the 1986 legislation, thus immunizing the regulation from the taxpayer's attack on its validity.'

The court also reached the same conclusion as the Treasury in its analysis of the statute. The court read the statute as making grandfather protection available when the family cannot escape from the terms of transfer contained in the trust.47 Because Mrs. Peterson could exercise her power of appointment to undo the transfer made by her husband, the court determined that the statute required denial of grandfather protection.

In Simpson v. United States,49 the facts were similar to those in Peterson. Unlike Mrs. Peterson, however, Mrs. Simpson did not allow her general power of appointment to lapse. Instead, she exercised it in favor of her grandchildren, which resulted in an outright distribution of the trust corpus to them upon her death.50 Not surprisingly, the Service relied heavily on Peterson. The Simpson court, however, found the factual differences between the cases warranted a different outcome given the language of the then-temporary constructive-addition regulation.51 Under the regulation, a constructive addition occurred "where any portion of a trust remains in the trust after the release, exercise, or lapse of a [general] power of appointment. . . .."52 Whereas in Peterson, the trust corpus remained in the trust after the lapse of Mrs. Peterson's power, none of the corpus remained in the Simpson trust after Mrs. Simpson exercised her power.53 In other words, as perceived by the Simpson court, the critical condition in the regulation, which was satisfied in Peterson but not in Simpson, was the requirement that the corpus remain in the trust after the exercise or lapse of the power.54

The Simpson court's analysis of the regulation reveals a weakness in its drafting. Under the Simpson court's reading, in order to defeat grandfather protection, there must be a release, exercise, or lapse of a general power of appointment;55 and trust corpus thereafter remaining in the trust 56 Although the text of the regulation easily lent itself to this reading, the 1980 preamble strongly suggested that the Treasury intended otherwise. The preamble implied that the exercise, release, or lapse of a general power-being a taxable event under either the estate or gift tax-would be sufficient to defeat grandfather protection (without regard to whether the trust corpus remained in the trust57).58

What accounts for this apparent conflict between the regulation and the preamble? The preamble revealed the Treasury's policy-based impulse to make the availability of grandfather protection turn on taxability under the estate and gift tax.59 The preamble explicitly offered two policy justifications for making such taxability determinative. First, where the corpus of a trust becomes the subject of a taxable event, the beneficiaries can alter or undo the terms of the initial transfer.60 Second, introducing a rule that failed to parallel estate and gift tax principles would be inappropriate.61 The regulation's text, on the other hand, disclosed a language-based concern that the statute might not be able to accommodate a construction focusing solely on taxability, because, ordinarily, the word "addition" contemplates an increase or augmentation.62

The Treasury apparently sought to resolve this conflict between the regulations and preamble by adopting a fiction: if the corpus remained in the trust after the exercise, release, or lapse, it could be viewed as if the trustee distributed the corpus to the beneficiaries who then recontributed the corpus to the trust, thus effecting an increase in the trust corpus 63 This fiction, however, has its difficulties. Consider the difference between a lapse and an exercise. An exercise ordinarily entails a direction that the trustee make immediate distribution.64 In contrast, a lapse typically results in the trustee retaining the corpus.65 While invoking the fiction in the case of a lapse is relatively easy (given that after the lapse, the corpus remains in the hands of the trustee), how to apply the fiction when the power is exercised and, as a result, the corpus immediately leaves the trustee's hands is unclear.

Although the text of the regulation clearly contemplated that a constructive addition could occur in connection with an exercise, the only relevant example contained in the regulations involved a lapse.66 The drafters provided no guidance with regard to how the constructive-addition concept might apply to an exercise. Faced with this lacuna, the Simpson court distinguished Peterson as a lapse case and upheld the grandfather protection based on its reading of the temporary regulation's text.67

The court's decision in Simpson, however, ultimately is not satisfying. The court fails to reconcile its reading of the regulation with the language in the regulation indicating that an exercise could create a constructive addition. The court also fails to explain the inconsistency between its reading of the regulation and the preamble. In addition, in emphasizing the text of the regulation, the court fails to appreciate the consequences in policy terms. The differentiation between an exercise and a lapse, a distinction long ago abandoned in the estate and gift tax context, makes no sense.' Put simply, it seems wrong to tax the Peterson family more harshly than the Simpson family merely because Mrs. Peterson's advisers recommended that she allow her power to lapse in favor of her grandchildren rather than exercise it.

The Simpson court was not satisfied with its regulation-based distinction, concluding that the statute itself clearly called for the result the court reached.69 Although the Simpson court did not acknowledge the conflict created with the Peterson decision, the court did reject implicitly Peterson's holding that the question of grandfather protection should turn on whether a taxable transfer under the estate and gift tax provisions has occurred.70 Contrary to the Simpson court's holding, under the Peterson court's taxable-transfer analysis, Mrs. Simpson's exercise of her general power of appointment, a taxable event for estate tax purposes,71 would have precluded grandfather protection.

C. New Regulations

Obviously disappointed with the defeat in Simpson, the Treasury issued new regulations overruling Simpson and reaffirming Peterson.72 In the preamble to the proposed regulations, the Treasury explicitly discussed the conflict between these decisions and stated its preference for Peterson.73 Surprisingly, however, the regulations do not remain faithful to the taxable-transfer theory; the Treasury did not even acknowledge that the Peterson court had concluded that the theory, as adopted by the Treasury in 1980 and later endorsed by Congress' ratification in 1986, "would ensure that the GST was given the scope intended by Congress."74 In the preamble to the final regulations, the Treasury rejected commentators' suggestions to make the regulations consistent with the theory.75 As will now be discussed, the Treasury's decision to disavow the theory is crucial in terms of another aspect of grandfather protection addressed in these regulations.

The other aspect of grandfather protection deals with a technique developed after the 1986 legislation to stretch out the term of grandfathered trusts. The technique is part of an emerging pattern: lawyers seeking to achieve their clients' tax-planning objectives by prevailing upon state legislatures to enact provisions that will facilitate federal tax savings.76 This technique requires state legislation that authorizes trustees who have a power to invade corpus to make the invasion in further trust.

As previously indicated, when Congress enacted the 1986 version of the GST, it embraced the provision in the regulations under the 1976 legislation applicable to powers of appointment. Thus, subject to the perpetuities limitation already discussed, a donee of a power could extend a pre1986 trust containing a special power of appointment, without causing a loss of grandfather protection, if the donee of the power exercised it by creating a new trust with a longer duration. As a result, estate planners began to suggest that such powers be exercised in this manner in order to maximize transfer-tax savings.77

However, when a pre-existing trust did not contain a power contemplating the creation of a new trust, how the trust's term could be extended was not immediately clear. Ultimately, the estate planning community fashioned a solution: secure state legislation authorizing trustees with a power to invade under the instrument to make the invasion in further trust, even if the instrument did not explicitly contemplate the creation of a new trust.78 Because a trustee's invasion power is a type of special power,79 the theory was that trustees could, under such state legislation, exercise the power by creating new trusts with extended terms without losing grandfather protection (provided that no violation of the federalized perpetuities provision occurred).80 This legislation would have a retroactive character if it was to have the desired effect: it would apply to trusts previously created.

Understandably, the Treasury found this technique offensive and adopted an approach in the new regulations designed to close it down.81 Under the regulations, the distribution to a new trust constitutes a modification that could result in a forfeiture of grandfather protection:82 modifications defeat grandfather protection if vesting is deferred or any beneficial interest is shifted to a lower generation.83 In other words, the regulations will not tolerate a tax-saving invasion in further trust (i.e., an invasion having the effect of prolonging grandfather protection). However, this new modification rule does not apply if the original trust or state law in effect at the time the trust became irrevocable explicitly authorized the trustee to make distribution in further trust without the consent of the beneficiaries or the courts (though the terms of the new trust cannot violate the federalized perpetuities rule and still retain grandfather protection).85

Parenthetically, the regulations impose a rather draconian penalty where an invasion in further trust is treated as a disqualifying modification. The new trust is not entitled to grandfather protection, and the old trust loses its protection as well. The loss of grandfather protection is triggered even if the new trust is created by a partial invasion of, say, one dollar of the old trust's corpus. By creating such a penalty, Treasury has created an in terrorem effect. Taxpayers inclined to make such an invasion will be intimidated from doing so once they realize that, if the regulations are sustained, it will result in the loss of all grandfather protection.

The Treasury could not continue to adhere to the taxable-transfer theory while, at the same time, closing down the invasion-in-further-trust technique. Under the taxable-transfer theory, a trustee's invasion in further trust, not a taxable event for estate or gift tax purposes,87 would not defeat grandfather protection. Moreover, given the policy justification the Treasury proffered when it originally adopted the theory in the regulations under the 1976 legislation-that grandfather protection should remain intact if the family cannot alter or undo the plans88--the technique should not result in a loss of grandfather protection because the family cannot overrule a trustee's decision to create a new trust with a longer term if state law, whenever enacted, authorizes it. Thus, to eliminate the technique, the Treasury needed to abandon the theory, as well as its policy justification.89

In doing so, the Treasury to a great extent ignored the terrain it had in large part created. First, although embracing in the preamble to the proposed regulations the aspect of the Peterson court's analysis that relies on the taxable-transfer theory's justification.' the Treasury nevertheless rejects the theory itself in the preamble to the final regulations 9' Second, although the Treasury continues to endorse the policy justification in the power-of-appointment setting, it refuses to permit the justification to operate in the context of the invasion technique. Third, the decision to abandon the theory came after the government had convinced the Peterson court that Congress had ratified the theory in 1986. And finally, the Treasury overruled the Simpson decision in the new regulations, even though the court held that the statute itself clearly mandates its holding.92 In effect, rather than continuing to litigate with taxpayers, the government declared victory by regulation, choosing to uphold the aspect of the Peterson decision that it found favorable, rejecting the aspect of the Peterson decision that it found unfavorable, and completely disregarding the Simpson decision.

At one time, the courts would have frowned upon such opportunistic rewriting of regulations, but the Supreme Court's Chevron doctrine has had a transformative impact on the power of the agencies, including the Treasury.93 Part II will examine the resulting expansion in agency authority, and Part III will return to a consideration of these regulations, focusing on their validity in the face of Congress' ratification of a contrary approach in 1986 and the evolving Chevron doctrine.

PART II. THE CHEVRON DOCTRINE

In recent years, under its Chevron doctrine, the Supreme Court has preferred that agencies (including the Treasury), rather than the courts, resolve questions of statutory ambiguity through the writing, and rewriting, of regulations. Even prior to Chevron, the Court required deferential review of agency interpretations,95 but Chevron made two changes.96 First, it increased the level of deference.97 Previously, the courts were required to examine a variety of factors in determining whether or not in any given case the interpretation was persuasive and, therefore, entitled to deference 98 Now, Chevron requires courts to give controlling deference to an interpretation if the statute is ambiguous and the agency's approach constitutes a reasonable resolution of that ambiguity.

Second, Chevron began to justify deference in a new way. No longer focusing exclusively on agency expertise as a justification, the Court began to emphasize political accountability.100 As a part of the executive branch, agencies are more politically accountable than courts and are therefore a more suitable repository for interpretive responsibility.101 The connection between political accountability and interpretive responsibility flows from the growing realization that law construction is often the equivalent of lawmaking.102 Law construction entails the making of policy, a function better served, under the Court's new theory, by the politically accountable agencies rather than by the politically insulated courts.103 Thus, under Chevron, deference no longer rests solely on agency expertise for its justification but rather, as the Court stressed, on an agency's political accountability as well.104

This is not to suggest, however, that in the absence of deference, interpretive decisions would be politically unconstrained. On the contrary, if a judicial interpretation proved to be problematic, Congress obviously would be urged to overturn it. And, to the extent that legislative redress were not provided, Congress-as well as the President-would face political consequences. Thus, in a nondeference world the judge would have more decision-making discretion, but political accountability would nevertheless continue to exert an important influence.15 The question, in other words, is not whether political accountability should have a role in the interpretation of statutes, but rather what king of role it should play.

Chevron's reach in the tax area had, until recently, been the subject of some question.'01 It was at first unclear whether Chevron applied to the Treasury's interpretive regulations, because unlike legislative regulations, there is no statutory requirement that the public be given notice and an opportunity for comment before interpretive regulations are promul

gated. Even though the Treasury in fact provides notice and opportunity for comment in issuing interpretive regulations, some suggested that Chevron might not apply in this context.' Second, given the specialized nature of the Tax Court and its expertise, there was some doubt whether it would be appropriate to impose Chevron and thereby require the court to defer to the expertise of the Treasury."' Both of these questions, however, are now be resolved. The Supreme Court has made clear that Chevron's framework is applicable to interpretive regulations"' and that the Tax Court, despite its own expertise, is required to give Chevron deference to the Treasury's regulations."' Recently, moreover, the Supreme Court held that an interpretation issued without notice and comment might nevertheless be entitled to Chevron deference.112

Under Chevron, once a court finds that a statute is ambiguous, the court is obliged to defer to any reasonable resolution of the ambiguity embodied in a regulation-provided Congress contemplated that the agency would have interpretive authority and the agency issues its interpretation in a format Congress anticipated would be binding on the courts."' Thus, in any case involving the construction of a statute under which an agency has interpretive authority,"' the threshold question is whether the statute is ambiguous. "' The lower courts' disagreement about the meaning of a statute may suggest that the statute contains the requisite ambiguity to invoke Chevron deference. In Smiley v. Citibank,"6 the Court, in making the threshold inquiry, emphasized that the different readings the statute had received in the Supreme Courts of New Jersey and California was in itself a strong indication of ambiguity."' If courts apply this approach at the federal level-and no apparent reason exists why disagreement in the federal courts should be viewed differently-Chevron may have transformed the role of the Supreme Court itself.' " Inter-circuit conflict traditionally has been a basis for granting review of tax litigation in the Supreme Court, but such conflict may now argue in favor of deference to the Treasury's construction, resulting, as a practical matter, in a limit on the Court's authority." 119

Indeed, it is arguable that the Court uses its resources improvidently when it grants review in a tax case to resolve an inter-circuit conflict, or, for that matter, whenever the statute is ambiguous. Under Chevron's logic, the Court should yield to any post-decision regulation that reasonably resolves the ambiguity, even if that regulation is contrary to the approach adopted by the Court. Although permitting an agency to replace the Court's interpretation of a statute with its own contrary interpretation is novel and inconsistent with the traditional understanding of the role of the judiciary," Chevron's preference for agency resolution of statutory ambiguity does lead in this direction. 121 In Estate of Hubert v. Commissioner,122 though not citing Chevron, the three-justice concurring opinion followed this lead and invited new regulations incorporating the very argument that the government had advanced unsuccessfully before the Court. 123 Shortly after the Hubert decision, the Treasury accepted the invitation and overturned the Court's decision by issuing regulations.124 From this vantage point, it would seem that the Court made an unwise commitment of resources in deciding to grant review in Hubert. 125

Chevron is potentially transformative in another important way. Given its political-accountability underpinnings, agency-administered statutes may no longer have a fixed meaning."6 In Smiley 127 a regulation was promulgated approximately one hundred years after the enactment of the underlying statute.128 After acknowledging the traditional view that a regulation issued contemporaneously with the enactment of the statute ordinarily receives deference on that account, the Court in Smiley concluded that the delay was of no consequence.129 The Court reasoned that because Congress presumedly intended for ambiguities to be resolved by the politically-accountable agencies, the validity of a regulation is not undermined by a lapse in time.130

Because contemporaneousness is no longer necessary, the question becomes whether agencies must remain consistent. If, for example, an agency adopts one construction of an ambiguous statute, can it later adopt the opposite construction if a president with a different political philosophy is elected even though there has been no intervening legislation? Emphasizing again the political-accountability theory behind Chevron, the Court indicated that, in general, consistency is not required and that, absent an abuse of discretion or a failure to respect reliance interests, the agencies generally remain free to amend their regulations and alter positions previously taken."' Though agencies remain required to proffer a satisfactory justification for a change in position (imposing on agencies a duty to be consistent in a way that is somewhat analogous to the duty that the doctrine of stare decisis imposes on the courts)," Chevron has led to a greater willingness to embrace such change.133

Indeed, following Chevron's political-accountability theory to its logical conclusion, one might argue that whatever remains of the agencies' obligation to justify change ought to be eliminated. Consider, for example, the propensity to issue regulations in the waning hours of the administration. No longer facing the same intensity of political consequence, a lame duck administration understandably would be tempted to issue regulations that would have been politically unappealing at an earlier point in its tenure.134 An incoming administration should to be able to reverse course without needing to supply a justification, given the friction of political consequences that will ensue and the diminished level of friction facing the prior administration when it adopted the interpretation. Stated differently, why should an agency decision, once made, be viewed as presumptively valid and binding if its legitimacy rests on the erroneous premise that it was the product of full, active political calculation?135 Moreover, even when the agency's initial position is forged under the pressure of fullblown political forces, it seems inconsistent with Chevron to give the courts the gate-keeping function of deciding whether the change is justified because the decision to deviate from the initial position is equally informed by political considerations.

In short, with delay irrelevant and consistency not essential, agency-- administered statutes containing ambiguities become mutable"6--or, to borrow from the constitutional lexicon,"living documents"37--no longer having the meaning fixed by Congress"' at the time of their enactment."9

Chevron's implications, as embellished by Smiley, may presage the end of an inveterate aspect of tax litigation. In the past, the government's defeat in Simpson likely would have led to further review in other circuits, or perhaps in the Supreme Court, on the ground that it conflicted with Peterson. Now, the government instead simply writes a new regulation announcing the result it failed to secure in court. As indicated, rather than becoming a predicate for Supreme Court review, inter-circuit conflict becomes evidence of statutory ambiguity, making the Treasury, not the Supreme Court, the ultimate interpretive authority. If the tax bar at one time viewed the Treasury as a mere adversary, that view no longer accurately reflects the more dynamic role the Treasury now enjoys. In short, given its enhanced quasi-legislative function under Chevron, the government is no ordinary adversary because it can rewrite the rules in many cases rather than litigate the meaning of the rules as originally written.

Is this a salutary alteration? The answer is not clear. On the one hand, allowing the Treasury more influence is valuable because of its enormous expertise-an expertise understandably lacking in many judges sitting on tax cases." Unlike the courts, the Treasury is able to bring this expertise to bear on an entire area of law at one time, facilitating an appreciation of the various ways in which the rules it promulgates interface. Also, Congress may not be able to respond as quickly as the Treasury to resolve issues not contemplated at the time of the statute's enactment.141 Moreover, Congress may be completely disabled from acting because nonpolicy-based concerns trump any legitimate policy objective.112 And, as some commentators have suggested, increased deference tends to create more uniform application of the law by reducing the potential for disagreement among the circuit courts.143

On the other hand, there is the question of the Treasury's bias.144 It would, for example, be difficult to maintain that the government's loss in Simpson did not, in some way, affect its perspective. Indeed, the Treasury's very position as the taxpayer's adversary in tax litigation will tend to produce bias. Just as criminal prosecutors are not given the quasi-legislative responsibility of defining the elements of the crimes they prosecute, so too, one might argue, more skepticism would be appropriate regarding the scope of the Treasury's lawmaking function.145 Although judges are certainly not free of bias,'"4 at least they do not suffer the bias one acquires as an adversary."' Thus, if disinterested, unbiased analysis is the objective,'" one can make a fairly compelling argument that Chevron's shift of power from the courts to the agencies"49 is not entirely desirable.`So

One might also negatively view this alteration because of the resulting diminution in the courts' authority to limit the abusive exercise of power by another branch of government."' The Service has recently been perceived as an unresponsive bureaucracy. `52 To the extent that a disinterested judge might be able to restrain bureaucratic power, Chevron can be seen as bureaucracy entrenching. This is somewhat ironic. As the perception of the Service has grown more negative, a corresponding popular impulse to curtail its authority has arisen. 113 Oddly, at the very time this impulse took root, the courts enhanced the government's authority through Chevron in the name of political accountability. In other words, the Supreme Court has, in effect, enhanced the power of an unpopular agency in the name of sensitivity to popular will.

Finally, although Chevron's political accountability justification has been critiqued elsewhere,"4 a few negative critical comments follow. First, the Court's recent decision in FDA v. Brown Bc Williamson Tobacco Corporation'55 raises a new question about the justification. In Brown & Williamson, the Court took the view that it would be inappropriate to infer that Congress had intended that the FDA could use its interpretive authority to decide whether to regulate tobacco.?56 The Court reasoned that if Congress had contemplated that the FDA could decide such an important issue, it would have said so affirmatively and clearly, rather than remain silent and assume the decision could be made under the agency's general interpretive authority.157

In thus limiting the agency's discretion, the Court appears to have engrafted a significant qualification onto Chevron. "8 Under this qualification, the more important the issue, the less likely the agency's decision will be entitled to deference. This qualification seems contrary to the very premise of political accountability: that the resolution of issues by the agencies will produce a political response, and the agencies, anticipating such a response, will either behave appropriately or pay a political price. If the issue is an important and visible one, the qualification renders Chevron deference unavailable. If, on the other hand, the issue is unimportant and not very visible, the agency will receive deference; however, in this latter context, the probability of political consequence is greatly diminished. In short, as so qualified, Chevron produces deference where political consequence is least likely, a result completely contrary to its political-- accountability theory.159

Second, the current controversy about campaign finance is also relevant to the merits of the justification. As many have argued, in the absence of reform, our system of democracy will remain in disrepair,160 with campaign contributors influencing elected officials, who in turn, seek to influence (and, in the case of the President, direct) the decisions of administrators.161 As long as campaign contributions are permitted to have such a negative impact on the decision-making process, the system will be better served by an approach that gives more authority to politically insulated judges than to politically vulnerable administrators.162

Third, institutional differences between the courts and the agencies must be considered in assessing the justification. It would seem that an underlying premise for the justification is the assumption that the agencies and the courts, when called upon to interpret ambiguous statutes, engage in approximately the same type of activity. If this is a valid premise, the balance of Chevron's logic is inexorable. Because resolution of statutory uncertainty by a court or an agency is essentially the same activity, it is preferable that the more politically accountable of the two be assigned this responsibility.

However, the ways that agencies and the courts execute their interpretive function differ. First, the agencies have more expertise as well as a superior ability to engage in the kind of cost-benefit analysis that legislators make.163 Second,judges are required to rely on canons of construction and view themselves as bound by prior decisions as a matter of stare decisis.164 While agencies are obviously not oblivious to these doctrines, they remain largely unencumbered by the restrictions these doctrines impose.165 Third, some suggest that judicial restraint requires the judge to set forth explicitly the principle that animates each decision so that the judge's opportunity to indulge a personal policy preference in future cases is constrained." More generally, judges' decisions are the product of various nonpolitical incentives that narrow the range of available alternatives and lead to more principled interpretations.167 Agency decision-making, in contrast, tends to be less principled in this sense and more responsive to policy or political considerations. Also, as discussed, agencies are influenced by a kind of self-interest that does not impinge on the judge.

Fourth, though agencies ordinarily give their interpretations prospecfive effect,168 judges are required to make their decisions on a retroactive basis.169 Like the canons of construction and stare decisis, retroactivity produces more consistent decision-making over time and concomitantly limits the range of alternatives available to the judge.170 Fifth, unlike the agencies, judges are not permitted to act in the absence of a case or controversy.171 Sixth, judges, unlike the agencies, are required to decide cases incrementally, resolving only the issue presented and leaving for future determination the larger issues that need not necessarily be addressed immediately.172 Both the case-or-controversy requirement and the incremental method tend to shape and limit the judge's perceptual field, thereby further constraining the scope of the judge's decisionmaking discretion."3

These differences cut two ways. On the one hand, the wider scope of discretion that the agencies enjoy, as well as their enhanced ability to make legislative-type judgments, creates more flexibility and, ultimately, efficiency in the face of changing circumstances." On the other hand, the constraints under which the courts operate are likely to reduce the influence of extraneous considerations such as adversarial bias or other forms of institutional self-interest that may undermine the objectivity of some agencies. In short, the premise that agencies and courts engage in the same kind of interpretive activity does not withstand comparative analysis. Perhaps it would be more accurate to say that there is an efficiency-objectivity tradeoff, and that Chevron's implicit choice of efficiency over objectivity due to political accountability is not without its cost.

PART III. THE REENACTMENT DOCTRINE AND ITS RELATIONSHIP TO CHEVRON

As discussed in Part I, the government did not issue the new GST regulations until after it had first persuaded the Second Circuit in Peterson"' that Congress ratified the version of the regulations issued under the 1976 legislation when it reenacted the GST in 1986.176 Indeed, in its brief, the government argued that the earlier "regulation received an overall endorsement" from Congress.177 And, in upholding the regulations, the court specifically referenced the reenactment doctrine.8

Generally, when the courts invoke the reenactment doctrine, they view an interpretation as authoritative-as if Congress had enacted the interpretation-based on Congress' reenactment of the legislation under which it was originally issued." Yet, even though Congress reenacted the GST in 1986 and, in doing so, explicitly approved of the earlier regulations, the Treasury has made the new regulations less friendly to taxpayers.180 Part III will focus on the relationship between the reenactment doctrine and Chevron, ultimately raising the question of whether the new regulations are valid given the reenactment doctrine.

As suggested in Part II, Chevron enhances agency authority in two ways. First, Chevron requires the courts to defer to agency interpretations when the statute is ambiguous and the interpretation is a reasonable one. Second, it contemplates that agencies will find it necessary to change their interpretations over time as circumstances evolve and such change, even if brought about as a result of a change in administrations, may be entitled to deferential review as well. 181

What is the relationship between Chevron and the reenactment doctrine? To the extent that the reenactment doctrine is available to the government as an argument in support of an agency interpretation, the doctrine seems to be consistent with Chevron's pro-agency stance. On the other hand, to the extent that the doctrine is used to prevent an agency from changing an interpretation, it conflicts with Chevron's willingness to embrace change.

A. Reenactment Doctrine Invoked By the Government

Consider first the relationship between Chevron and the use of the reenactment doctrine when the government invokes it. If the original statute is ambiguous, its reenactment makes a modest contribution, if any, to the argument in favor of the interpretation, for the ambiguity itself makes Chevron deference appropriate and thereby leads to a validation of the interpretation (although the reenactment doctrine might nevertheless have some significance in Chevron's second step, i.e., in determining whether the interpretation reasonably resolves the statutory ambiguity).182 In other words, although application of the doctrine is consistent with Chevron in that it results in validating the agency's interpretation,"83 Chev ron alone would produce that outcome without any assistance from the reenactment doctrine."' In contrast, when the original and reenacting statutes are unambiguous and cannot accommodate the agency's interpretation, the doctrine may have more importance. On the one hand, Chevron seems to preclude a court from utilizing the doctrine to uphold an interpretation contrary to unambiguous statutory language."' On the other hand, a court might be reluctant to invalidate an interpretation when Congress reenacted the statute after having been made aware of the interpretation, irrespective of the clarity of the statutory language.116 On the whole, the reenactment doctrine as a government argument in support of an interpretation has rather meager utility when the statute is ambiguous, but perhaps is more useful when the statute is unambiguous.

B. Reenactment Doctrine Invoked Against the Government

Consider now the use of the reenactment doctrine against the government. If an agency issues an interpretation under the original statute but then seeks to change its position after reenactment of the statute, the doctrine, if successfully invoked, will prevent the agency from making the change. Precluding an agency from reconsidering its position in this way constitutes a significant constraint on agency discretion and denies the agency the ability to respond to evolving circumstances.' However, agency discretion is no less constrained when Congress incorporates its scheme in unambiguous statutory language.

When the original and reenacting statutes are ambiguous and would both accommodate the pre-reenactment interpretation, one could make a plausible case for invoking the doctrine against the government. "I Or even when the statutory language unambiguously conflicts with the interpretation, a case could possibly be made for using the doctrine if there is clear evidence that Congress endorsed the interpretation.189 After all, if Con gress was aware of the interpretation and found it was inconsistent with the reenacting statute, presumably Congress would not have adopted the new legislation in silence, but would have disclosed its discomfort with the interpretation. However, one could also argue that Congress' silence indicates Congressional intent to ratify the interpretation, but, at the same time, leave the agency with discretion to make modifications. Yet another reading that would support a rejection of the reenactment doctrine is that Congress was unaware of the statutory interpretation, suggesting that Congress intended the agency retain its ordinary discretion under Chevron to resolve any ambiguity in the reenacting legislation. Ultimately, whenever the doctrine is invoked against the government, the question must be whether Congress intended not merely to embrace the interpretation, but also to preclude the agency from making any post-enactment alterations."

This, in turn, leads to the normative question of how to fashion a default rule. How, in other words, should the courts presumptively interpret legislation where Congress fails to make its intent explicit with regard to a pre-reenactment interpretation?"' On the one hand, the flexibility made available when agencies are permitted to make changes as circumstances dictate suggests it would be more efficient to apply a default rule that allows interpretations to be modified after reenactment. Also, from a public choice perspective,192 one can argue that if Congress fails to freeze an interpretation in place, permitting the agency to alter its position is preferable because rewarding Congress for its ambiguity is inappropriate or even counterproductive.193 On the other hand, the concern that agencies might alter their interpretations opportunistically supports a default rule that would preclude post-reenactment modifications.94

The development of the law on this issue has not been linear. Long before its decision in Chevron, the Court, alluding to a concern about constraining agency discretion, concluded that the doctrine could not be invoked against the government.?5 However, several years before its decision in Chevron, the Court retreated and applied the doctrine after finding in the legislative history accompanying the reenacting legislation an intent to freeze in place an agency interpretation." Then, shortly after Chevron, the Court in Motor Vehicle Manufacturers Assn v. State Farm Mutual Automobile Insurance Co.,"97 qualified its approach yet again. Emphasizing Chevron's willingness to permit agencies to deviate from earlier interpretations, the Court distinguished between use of the doctrine by the government and against the government and indicated that it would be more demanding of litigants seeking to invoke it in the latter context."8

The Court's retreat is not surprising. After all, Chevron effected an expansion in agency authority not only by creating greater latitude for agencies to change their interpretations, but also by granting increased deference. The reenactment doctrine, when applied against the government, potentially conflicts with both of these Chevron strands. The spirit of the pro-change strand is undermined if courts apply the doctrine to prevent the agency from reconsidering its original position. Likewise, the spirit of the pro-deference strand is frustrated if the agency is so constrained. In short, although this tension between Chevron and the reenactment doctrine requires further clarification, one can safely conclude that ardent proponents of Chevron will be strongly inclined to resolve it by requiring rather clear evidence of anti-modification intent as a predicate for invoking the doctrine against the government.'"

Recently, in FDA v. Brown & Williamson Tobacco Corp. ,2 the Supreme Court invoked the reenactment doctrine against the government, freezing in place a prior agency interpretation. The tension between the doctrine and Chevron surfaced in the interaction between the majority and dissenting opinions, with the majority permitting the doctrine to trump Chevron's pro-change principle and the dissent choosing Chevron over the doctrine. The majority applied the reenactment doctrine against the government in a relatively uncomplicated context and, as will be suggested, did so gratuitously. Moreover, the Court's analysis sheds little light on the kind of evidence Congress must provide before a court is justified in finding a freeze was intended.

In Brown & Williamson, the FDA originally had interpreted its governing statute as precluding the agency from regulating tobacco. Subsequently, Congress enacted various pieces of legislation regulating the sale of tobacco. The FDA then abandoned its original interpretation and took the position that its governing statute conferred upon it the necessary regulatory authority. The issue before the Court was whether the FDA exceeded its authority in asserting regulatory jurisdiction. Both the majority" and the dissent?' appeared to be in agreement as to the proper framework for determining the validity of the FDA's new interpretation: if the assertion of jurisdiction was contrary to the statute's unambiguous mandate, Chevron would require that the interpretation be invalidated.

The majority then examined the FDA's governing statute and the tobacco-regulating legislation Congress enacted over the years. If the governing statute was construed as permitting the assertion of jurisdiction, the majority reasoned that the FDA necessarily would be required to ban the sale of tobacco entirely."' Yet, the majority found that the tobacco-- regulating legislation, enacted subsequent to the governing statute, clearly and unambiguously contemplated the continued sale of tobacco.' Thus, the FDA's assertion of jurisdiction and the complete ban that would necessarily result could not be sustained.

Given the majority's conclusion in Brown & Williamson that there was a conflict between the FDA's assertion of jurisdiction and the clear tobacco-regulating legislation, one might have anticipated that a reference to the Chevron notion that an agency cannot override a clear statute would suffice. However, the Brown & Williamson majority expanded its analysis to include the reenactment doctrine. The majority observed that Congress had enacted the tobacco-regulating legislation against the backdrop of the FDA's original interpretation that it did not have jurisdiction. Relying on one of its reenactment decisions, Bob Jones University. v. United States,25 the majority concluded that Congress intended to ratify the FDA's original interpretation and, in effect, freeze it in place.206 However, this use of the doctrine was gratuitous. The tobacco-regulating legislation, the majority concluded, unambiguously contemplated that the FDA would not have jurisdiction. Under the Chevron framework, an agency interpretation in conflict with such unambiguous legislation cannot be permitted to stand. Thus, the legislation should have been viewed as a bar to the FDA's assertion of jurisdiction, making any examination of the prior interpretation or the reenactment doctrine unnecessary.

It is difficult to infer from the Brown & Williamson majority's willingness to apply the reenactment doctrine against the FDA the circumstances in which the Court will be prepared to invoke the doctrine against the government in future cases. After all, given the majority's view of the tobacco-regulating legislation, the case was a relatively easy one in which to examine the relationship between Chevron and the reenactment doctrine. Once the Court found the legislation clearly precluded the assertion of jurisdiction, the agency's authority to deviate from its original interpretation could not be sustained, thus making for an easy, as well as gratuitous, application of the doctrine. In contrast, when the reenacting legislation is ambiguous, the analysis is more complicated. The Court, however, did not address the issues that arise with regard to an ambiguous reenactment: whether the doctrine is available against the government in this context; and, if so, how clearly Congress must express its intent in order to achieve this outcome. Because the majority in Brown & Williamson did not consider these issues, it failed to address an important distinction. In invoking the reenactment doctrine, the majority relied on Bob Jones. However, in Bob Jones the government invoked the doctrine to validate an agency interpretation, not against the government to constrain an agency's authority to modify an interpretation.208 Did the Brown & Williamson majority intend to imply that the same level of evidence concerning Congress' intent is required to invoke the doctrine against the government as is required when the government invokes the doctrine? Or did the Brown & Williamson majority simply fail to appreciate that, unlike the case under consideration, Bob Jones involved the use of the reenactment doctrine to validate the agency's interpretation?

Whatever one might make of this, the Brown & Williamson majority certainly made clear that, even under Chevron, the Court can invoke the reenactment doctrine against the government in order to constrain change.209 The Brown & Williamson dissent, on the other hand, maintained that, given Chevron's pro-change principle, the Court should permit the FDA to alter its approach.210 The majority did not dispute the validity of the pro-change principle but instead implicitly concluded that it was inapplicable because of its determination that Congress had intended to ratify the FDA's original interpretation. Therefore, although the dissent apparently took the view that the pro-change principle preempted application of the reenactment doctrine against the government,211 the majority took the contrary view that the doctrine can be invoked against the government without violating the pro-change principle. Unfortunately, both the majority and the dissent failed to fully clarify their view of the relationship between the pro-change principle and the reenactment doctrine.

Curiously, neither the Brown & Williamson majority nor the dissent considered the Court's prior treatment of the applicability of the reenactment doctrine against the government. As previously suggested, in Motor Vehicle Manufacturers,212 the Court had staked out its position on the issue by indicating that Congress should not be understood as precluding an agency from changing its position merely because it has ratified an agency interpretation, unless, in the Court's words, the reenacting legislation effects a "statutory incorporation."213 Although the Court did not elaborate on the meaning of this phrase, it seems to contemplate a presumption in favor of permitting the agency to change its position in the absence of an affirmative indication by Congress of a contrary intent. However, in Davis v. US.,'15 a tax case decided six years after Motor Vehicle Manufacturers, the Court appeared more willing to apply the doctrine against the government.216 In Brown & Williamson, both the majority217 and the dissent,218 surprisingly, cited Motor Vehicle Manufacturers for the pro-change principle but neither opinion made any reference to the language presumptively disfavoring use of the doctrine against the government. Nor was any reference made to Davis or to the Court's earlier treatment of the issue.

How then should the Brown & Williamson majority be understood? One possible reading is that the majority intended to strengthen the reenactment doctrine as applied against the government. There are, however, alternative readings that would render the doctrine less expansive. After all, the Brown & Williamson Court found the legislation unambiguous,22 thus leaving unclear whether the Court is prepared to invoke the doctrine against the government in the context of an ambiguous reenacting statute. Perhaps when the issue arises again, the Court will seek to limit the majority opinion by claiming that the evidence that Congress intended to effect a freeze was overwhelming-tantamount to an incorporation. Or perhaps the Court will ultimately declare that in the majority's haste to bolster its analysis, it simply made a mistake in invoking the reenactment doctrine in the first place.

C. The Generation-Skipping Regulations and Reenactment

Thus, because of the distinction between the use of the reenactment doctrine by the government and against the government, the mere fact that the government successfully invoked it in Peterson221 does not necessarily lead to the conclusion that the earlier regulations thereby became unalterable. Nevertheless, it is conceivable that the courts will view Congress' explicit endorsement of the regulations as sufficient to freeze them in place, making the new regulations invalid. In any event, in normative terms, Congress' endorsement of the earlier regulations, even if viewed as falling short of the standard for applying the reenactment doctrine against the government, as well as the adversarial origin of the new regulations should certainly weigh against the validity of the changes. Under Chevron, however, these considerations are irrelevant.

After first examining the deference issue in the context of revenue rulings and ambiguous regulations in Parts IV and V, in Part VI the argument will be made that the validity of regulations should be assessed under the so-called Skidmore framework. This shift away from Chevron would make the courts less deferential and more circumspect, allowing various kinds of considerations to be taken into account that are out of bounds under Chevron.

PART IV. REVENUE RULINGS AND DEFERENCE

The transformation in the courts' deference to agency interpretations has not been limited to regulations. Traditionally, courts did not view revenue rulings as legally authoritative, and they received no deference, whether they were invoked by.. or against the government.224 A significant shift has occurred, however, in the post-Chevron years.225 Although the Supreme Court remains unwilling to commit itself on the issue,226 revenue rulings have been enjoying deference, with the level of deference beginning to crystallize.227 This is true not only when the government invokes a ruling. The courts have begun treating ruling-based arguments advanced by taxpayers with more deference as well.228 The effect of this alteration in the authoritativeness of revenue rulings is twofold: an expansion of the government's authority where it is the government invoking a ruling, and a contraction where it is the taxpayer.

In Davis v. United States.229 where the government invoked the revenue ruling, the Court gave it "considerable weight," but in doing so emphasized two elements: the ruling's longstanding status, and the reenactment of the statute after it was issued.' This, of course, left unclear whether these elements were crucial to the Court's decision to give the ruling such deference23 and whether Davis' deferential standard of review was the equivalent of Chevron deference was also left unclear.232

Five years later, although in the context of a ruling invoked by the taxpayer rather than the government, all of the Justices agreed, in Commissioner v. Schleier,233 that rulings were not to be viewed as having the same effect as a regulation." Although not entirely clear, this holding certainly suggested that a ruling was not entitled to Chevron deference (i.e., the controlling deference regulations enjoy). However, the dissent argued that rulings were entitled to "substantial" deference (citing Davis)' The majority, concluding that the ruling could not be sustained because it was contrary to the plain meaning of the statute,236 found no need to articulate the level of deference rulings were entitled to receive, and, therefore, did not respond to the dissent on this point.

In two recent decisions, Christensen237 and Mead,211 the Court made clear that courts should utilize a two-tier framework to determine the validity of agency interpretations 239 Under this framework, selecting which of two levels of deference is appropriate depends on whether Congress intended to permit the agency to exercise lawmaking responsibilities, and, if so, whether Congress intended that the responsibilities could be exercised through the particular type of interpretation issued by the agency." When a court concludes that Congress intended that the agency exercise lawmaking responsibility and that it have the authority to do so in the type of interpretation at issue, controlling (Chevron) deference is appropriate." However, when a court reaches the contrary conclusion, the interpretation is entitled to less deference under the so-called Skidmore standard.242

The Skidmore standard, under which tax regulations previously had been analyzed,243 requires courts to defer to a persuasive interpretation.' In determining whether the interpretation is persuasive (and therefore binding), the courts must make a contextualized judgment based on various factors that would be largely irrelevant under Chevron, including inter alia:11 the thoroughness of the agency's decision-making process, the consistency with which the agency has maintained its position, and the validity of the agency's reasoning (which, unlike the first two factors, is relevant under Chevron).' In applying persuasiveness as the standard, both Christensen and Mead could possibly be understood as suggesting that courts remain free to reject any interpretation with which they disagree. However, when these decisions are read against the backdrop of the language referenced in Skidmore, such an understanding becomes untenable. For example, consider a case in which the statute is ambiguous and would accommodate either of two constructions. If the agency adopted one of these constructions after thorough and logical analysis and then remained consistently faithful to it, Skidmore likely would require a court to defer to the agency's construction despite the Court's preference for the alternative.247

Whether the Court will definitively conclude that courts should analyze revenue rulings under Skidmore, rather than Chevron, remains to be seen. But given that all of the Justices agreed in Schleier that rulings do not enjoy the same authoritative effect as regulations (Chevron deference)" and that Christensen and Mead make clear that interpretations are entitled to either Chevron or Skidmore deference (with no intermediate level of deference), the Court likely will eventually clarify that Skidmore is the appropriate framework for analyzing the validity of rulings.249 If so, Davis ' "considerable weight" standard could be understood as a contextualized judgment about the ruling's persuasiveness, with its longstanding status and the statute's reenactment after Congress issued it explaining the heightened level of deference the Court applied."O

If Skidmore does become the governing framework, the lower courts may well need to reconsider the formulations they employ in evaluating the authoritative effect of rulings invoked by the government."' The circuit courts have been applying what might be called an intermediate level of deference, not as much as Chevron's controlling deference, but more than Skidmore's persuasive deference (perhaps reading Davis as imposing a per se rule of intermediate deference that is insensitive to context)?.112 Thus, if Skidmore is held applicable to rulings, a cutback in the level of deference will likely ensue in the circuit courts. Yet, it is also possible that any alteration in the deference formulation as a result of applying Skidmore might have little impact in these courts. Given the pressure of other matters, generalist judges lacking tax expertise may well find sufficient flexibility in the Skidmore standard to justify adopting the position taken in a ruling without engaging in a complete and independent analysis.253

The Tax Court, unlike the circuit courts, has adhered to the traditional view that rulings are entitled to no deference." Although reconsideration of this view would seem necessary in light of Christensen and Mead, the Tax Court is not likely to undertake this task eagerly. One way of understanding the court's no-deference stance is as a reflection of its own sense of expertise and, perhaps, "institutional ego." If this description of the court's approach is correct, the court presumably will resist taking Christensen and Mead seriously. The Tax Court could express this resistance in either of two ways: (1) by misreading Skidmore to require no deference whenever its own analysis leads to a conclusion that is inconsistent with the ruling, or (2) by adopting its own conclusion on the substantive issue and then framing it in terms that are compatible with the Skidmore standard. The court has already intimated an inclination to take the first of these two paths.

In those cases where the government is able to prevail because its ruling-based argument is reviewed deferentially, its authority is obviously expanded. Given the deference rulings enjoy (though perhaps a cutback will occur as a result of Christensen and Mead`), it would not be surprising if the government sought to expand its authority further by incorporating arguments it contemplates making in future litigation in a present ruling and then claiming the ruling as precedent.257 Moreover, as a practi cal matter, the government may frequently be able to secure controlling deference for its rulings, rather than mere Skidmore deference. For as discussed in Part V, the Supreme Court has made clear that an agency's interpretation resolving an ambiguity in its regulations is determinative "unless plainly erroneous or inconsistent with the regulation.258 Indeed, the Court recently deferred to the government's construction of an ambiguous regulation on the ground that longstanding rulings had adopted the same construction.259 Because so many tax disputes revolve around the meaning of the regulations, and since it is likely that, as a result of Chevron, an increasing number of issues will be addressed in the regulations, the potential for the government to dictate the outcome of litigation through rulings that resolve ambiguities in the regulations is quite considerable-any cutback that Christensen and Mead may precipitate notwithstanding. This expansion in the government's authority,260 though perhaps not as problematic as in the context of regulations because rulings will not universally enjoy the same level of deference as regulations, nevertheless raises the same kinds of questions about the failure to provide a disinterested check on governmental bias or bureaucratic abuse 261

The Supreme Court has not offered any more concrete guidance on the deference issue when the taxpayer invokes a ruling, but two decisions reveal a trend in the direction of binding the government by its rulings-a trend that is even more discernible in the circuit courts.262 In Commissioner of Internal Revenue v. Estate of Hubert,263 although none of the various opinions explicitly addressed the deference issue, the three-Justice concurring opinion predicated its rejection of the Service's argument on a concession made in a ruling with regard to the meaning of an ambiguous regulation.264 The two-Justice dissent did not disagree in terms of deference, but rather argued that the concurring opinion misread the scope of the concession.265 Thus, when the concurring and dissenting opinions in Hubert are read together, it appears that fives Justices may be prepared to take the view that rulings invoked by a taxpayer are authoritative and binding on the government266 (at least when the ruling clarifies the meaning of an ambiguous regulation267).

In Schleier, the Court concluded that a concession made in a ruling was not binding on the government because it was contrary to an unambiguous statute." Though not entirely clear, the negative implication is that the Court is prepared to hold a ruling binding on the government when the statute is ambiguous' and the ruling adopts a reasonable construction in light of that ambiguity.270 In any event, whatever one may make of Hubert's concurring and dissenting opinions and Schleier's negative implication, the Court apparently has established an outer boundary for deference and authoritativeness in this context: the government is not bound by a ruling if it can establish that the statute unambiguously calls for a contrary result.271

When a court concludes that the government is bound by a ruling, the government's authority obviously is contracted. Because taxpayers rely on rulings and because one important function that rulings serve is to facilitate transactions by making clear in advance the tax consequences they will generate, it is unfair272 and inefficient to permit the government to disavow a ruling after a taxpayer has consummated a transaction on the basis of the ruling.273 Thus, the contraction of the government's authority, resulting from the emerging view that rulings are authoritative and binding on the government, is constructive. While Hubert and the negative implication in Schleier are positive developments, the majority's unqualified assertion in Schleier that a court must disregard any ruling that is inconsistent with an unambiguous statute-apparently even in the face of taxpayer reliance274-- raises fairness and efficiency concerns. Indeed, in the aftermath of Schleier, the Service implicitly acknowledged its discomfort with the Court's decision to invalidate the ruling and thereby ignore the concession it contained, thus suggesting that the Service recognizes the salutary effect of standing behind its rulings.275

In sum, however the deference standard is formulated, revenue rulings now enjoy more deference than they did traditionally. The resulting expansion of government authority when a pro-government ruling is reviewed deferentially is problematic. On the other hand, when a pro-taxpayer ruling receives deference precluding the government from changing a position on which a taxpayer has relied, the constraint on government authority makes sense in terms of fairness and efficiency.

PART V. AUER DEFERENCE: AMBIGUOUS REGULATIONS

The deference transformation has been extended beyond the interpretation of statutes. The way that courts police agencies that interpret their own regulations has undergone change as well. Until recently, how deference applied in this context was somewhat unclear."" However, the Supreme Court has now clarified that a type of deference analogous to Chevron-type deference is appropriate. In Auer v. Robbins,277 the Court held that a court, when required to determine the meaning of an ambiguous regulation, must give controlling deference to an interpretation proffered by the agency if it is not plainly erroneous or inconsistent with the regulation.278

However, deference is not required if the interpretation does not represent the agency's considered view.279 For example, when the agency's appellate counsel argues in a brief that a court should construe a regulation in favor of the government, deference is inappropriate.' But when the brief reflects that the proffered interpretation represents the official view of the agency, and not simply the view of appellate counsel, the interpretation is entitled to controlling deference.281

The rationale is that, as long as the agency's interpretation is not inconsistent with the statute, any revised draft of the regulation containing the interpretation would necessarily be valid. It makes no sense, the argument goes, to reject as invalid an interpretation that could be validated so easily. The premise that the interpretation could in all cases be adopted by amendment is, however, open to some question. While, under Chevron, agencies are given great latitude to amend their regulations, this authority may nevertheless be subject to limitations beyond the requirement that the interpretation fall within the scope of the statute?" On the other hand, under Auer, the only limitations on the agencies' authority are the parameters of the statute under interpretation (as well as the requirement, of course, that the regulation be ambiguous and the government's resolution be reasonable)."

Two critical difficulties arise with what will presumably become known as Auer-type deference.285 First, under Auer, the problem of adversarial bias is most acute. There is little question but that the government's ability to determine objectively the most appropriate construction of a regulation will be less than optimal if the government must make the decision at the very moment of engaging in litigation with a taxpayer. Second, whenever a court upholds an agency's construction under Auer, the court implicitly applies that construction on a retroactive basis. In other words, while an agency might be able to amend an ambiguous regulation to incorporate a favored construction, that amendment would need to be adopted on a prospective basis if the agency did not have authority to adopt regulations retroactively.286 Thus, in the case of tax regulations, which can no longer be adopted on a retroactive basis,287 Auer permits the government to do indirectly what it cannot do directly.288

Another form of deference is closely related to Auer deference. When the meaning of a statute, rather than a regulation, is at issue and the agency argues in litigation that its interpretation should prevail, the question arises whether courts should defer to the agency's argument. Although the Supreme Court has not definitively resolved the issue, it appears to contemplate that an agency's litigating position is entitled to deference if: (1) the agency's interpretation represents the view of the agency (as opposed, for example, to the view of appellate counsel); and (2) the interpretation is supported by administrative practice.289

Assuming these two conditions are satisfied and, as a result, deference is appropriate, the question then becomes what level of deference should be applied. While further clarification of the issue is needed, the Court has implied strongly that an agency's litigating position should be analyzed under Skidmore." Thus, in contrast to regulations (which are entitled to

Chevron- or Auer-type deference), an agency's litigating position, while perhaps entitled to some deference, is not entitled to controlling deference. Again, the questions of adversarial bias and retroactivity are implicated, although they are less critical in this context because, under the Skidmore framework, the courts have more discretion to reject an agency's litigating position.291

PART VI. A PROPOSAL

The alteration in deference has had, and will continue to have, a very significant impact on tax practice, as well as on other areas of law. The Treasury's expanded regulation-writing authority under Chevron, which it used to overturn the Supreme Court's decision in Hubert.. and to resolve the Peterson/Simpson issue,293 indicates that the Treasury's role is undergoing a radical revision. The Supreme Court's decision in Auer also expands the government's influence by requiring that the courts defer to agencies' reasonable resolution of ambiguity in their regulations. At the same time, the increased deference given revenue rulings under the Skidmore framework294 contributes, as well, to the changing dynamic in tax litigation, by further strengthening the government's hand."

Whatever view one may take of the persuasiveness of Chevron's political accountability theory-and of deference in general-granting increased deference to the government raises two tax-specific issues. One relates to expertise, and the other to the government's position as an adversary in tax litigation.

First, deference ordinarily is justified not solely on the basis of political accountability, but also on the basis of agency expertise.' However, in the tax context, the government's expertise does not have the same importance as it does in other areas of the law. Because tax litigation often occurs in the Tax Court, which has its own substantial expertise, the danger that a decision will be either insensitive to relevant policies or the Code's structure is greatly diminished. In other words, to the extent that deference is driven by the concern that courts might otherwise undermine the agencies' expertise-based decisions, deference cannot be justified in areas of law when specialized courts are in place.

Under current law, however, the Tax Court does not have exclusive jurisdiction over tax cases. Although much tax litigation occurs in the Tax Court, generalist judges having less tax expertise than the government very frequently decide whether the government's interpretive position should be sustained. Taxpayers have the option of litigating tax cases in the District Court or the Court of Federal Claims,297 and the Courts of Appeals exercise appellate jurisdiction in all tax cases .. (with regard to questions of law) on a de novo basis." If this disparity in tax expertise between generalist judges and the government could be reduced, the case for increased deference in the tax area would be much diminished.

The question of how to reform the tax litigation process has been discussed elsewhere and is beyond the scope of this Article.? Nevertheless, two basic approaches that address this disparity in expertise are briefly considered. Under both approaches, one court, such as the Tax Court, would have exclusive jurisdiction over all tax cases.301 Where the two approaches differ is in terms of appellate review. One approach would create a new circuit court with jurisdiction to hear all appeals in tax matters.302 Under the other approach, as under current law, one of the existing circuit courts would conduct appellate review. But, because under the latter approach, generalist judges would review the work product of specialists-the lower court judges and tax administrators-limiting the scope of appellate review would be appropriate (otherwise, the disparity in expertise produced under current law would continue). Such a limitation on the scope of review could be achieved by requiring that the lower court's decision receive deference on questions of law as well as fact, making reversal on legal questions appropriate only where the ruling is unreasonable.303

Not only would these two approaches ameliorate the disparity-inexpertise problem, thereby weakening the case for increased deference, they also would create a salutary by-product. Under either approach, taxpayers would have less opportunity to engage in forum shopping; indeed, the former approach would completely eliminate forum shopping. In addition, the former approach would eliminate, and the latter approach diminish, the potential for non-uniform application of the law inherent in the current system (i.e., disagreements between the Tax Court and the circuit courts"'3" and disagreements among the various circuit courts).?s's

Second, the case for applying increased deference in the tax context is further undercut by another important distinction between tax and some other agency-administered areas of law. In tax cases, the government is the taxpayer's adversary. As the government's response to its defeat in Simpson suggests, adversarial bias is not easily put aside.' Therefore, although the government's litigation with a particular taxpayer has ended, the views formulated in the litigation take on a life of their own. In contrast, in other areas of law when the government is not a litigant and not suffering the burden of adversarial bias, the government can be trusted to make interpretive decisions that are not influenced by self-interest.

Consider the Supreme Court's treatment of the regulation in Smiley.307 The Court gave controlling (Chevron) deference to a regulation issued by the Comptroller of the Currency one hundred years after the enactment of the underlying statute and during the litigation that led to the Court's decision.308 Unlike the role the Treasury plays in the tax area, the Comptroller of the Currency was not a party to the litigation and, therefore, had no direct interest in its outcome.109 If the government were to issue a regulation resolving an issue in a pending tax litigation, permitting such a resolution to be binding under Chevron on the taxpayer would seem unjust. Indeed, Congress acknowledged as much by amending Code section 7805(b) in 1996 to eliminate the Treasury's authority to issue regulations on a retroactive basis.310 As a result of the amendment, the government is no longer permitted to determine the outcome of pending tax litigation by regulation.311 Nevertheless, subject to the proviso that the Treasury make any new regulation prospective, it remains free in many instances to rewrite the outcome of a decision it loses, as it did in the aftermath of its defeat in Simpson and Hubert.312 And, under Chevron, the government's adversarial bias is irrelevant in determining the validity of such a regulation. Moreover, Smiley suggests that, despite any statutory limitation on an agency's ability to make regulations retroactive, it would be inappropriate for a court interpreting a statute to give no deference to a regulation simply because it was issued after the consummation of the transaction under inquiry (though the Court presumably contemplates less than controlling deference in this context).313

Even more disconcerting, Auer requires the court to defer to an agency's interpretation when an ambiguous regulation is at issue.14 This has the effect of permitting the government to prevail whenever the litigation revolves around an ambiguous regulation-a common occurrence in the tax area. In contrast, when the government is not a party to the litigation but merely files an amicus brief offering its interpretation, Auer deference is obviously less troubling. Auer, in effect, permits the government to make its newly announced position applicable on a retroactive basis, in violation of the spirit of the 1996 amendment to section 7805(b).315 Not only does Auer contemplate that interpretations will be binding when issued after the transaction is consummated, it even contemplates that the interpretations will be binding when issued during the process of litigating the transaction's tax consequences. Permitting the government to declare its own victory when it appears as a party in the litigation undermines the effectiveness of the courts as a check on bureaucratic abuse and is fundamentally unfair.

In short, given the government's role in tax litigation, a compelling case can be made for reducing the level of deference it enjoys in tax matters. And should the disparity-in-expertise issue be resolved in one of the ways suggested, the case for reducing deference would become even stronger. The question, therefore, is how deference might be ratcheted downward in the tax area.316

In terms of regulations, whether interpretive or legislative, it would seem that Skidmore's framework would be preferable over Chevron's. Under Skidmore the court is permitted to consider a variety of factors in determining the validity of an agency interpretation,317 unlike the binding deference that Chevron requires.318 Thus, for example, a regulation issued long after the statute's enactment, or one having an adversarial origin, might be of questionable validity under Skidmore,"' although as Smiley indicates, such considerations are irrelevant under Chevron. 320 Or perhaps, the new generation-skipping regulations might be vulnerable under Skidmore because of Congress' ratification of the earlier regulation, even if it is assumed that the standard for applying the reenactment doctrine against the government has not been satisfied.

One could argue that, as an alternative to adopting the Skidmore framework, tax regulations should receive no deference. However, the difficulty with this approach is that the Code is complex and, as a consequence, there is a need for legislative-type rules that a court cannot easily supply. Given this institutional limitation, and given that Skidmore should provide sufficient protection against bureaucratic abuse, the no-deference alternative is not particularly appealing.12' Thus, Congress should enact legislation adopting the Skidmore framework as the controlling standard for evaluating tax regulations.

The deference required under Auer is perhaps most pernicious. The consequences of adversarial bias are at their worst in this context because the courts are, in essence, precluded from exercising any check on what amounts to the government's declaration of victory by retroactive regulation. Therefore, Congress should overrule Auer, eliminating all deference for the interpretations of ambiguous tax regulations proffered by the government during litigation 322 Similarly, Congress should overrule Smiley's conclusion that the courts must consult regulations even where issued after the transaction has been consummated and even if the agency does not have the authority to issue regulations retroactively.3

As indicated, while it appears that the government's litigation position regarding an ambiguous Code section's meaning does not receive as much deference as its interpretation of an ambiguous regulation,324 the same concerns about adversarial bias and retroactivity are implicated.325 Any deference-altering legislation should, therefore, also deny the government deference for its interpretations of ambiguous Code sections offered during litigation.326

Revenue rulings are more troubling. On the one hand, treating them differently from regulations is difficult to justify."' After all, both revenue rulings and regulations are the product of the government's careful deliberation.32$ It is, of course, true that regulations, unlike rulings, are issued only after the public has been given an opportunity for comment. Nevertheless, this distinction is not a satisfactory basis for requiring deference in one case but not in the other."' While the public-comment process entails some additional deliberation, the difference in the level of deliberation is not sufficiently substantial to warrant the discrimination. Furthermore, the government continues to consider its rulings after issuing them, qualifying, modifying, or revoking them based on the public's adverse reaction. This post-issuance deliberation of rulings appears roughly equivalent to the deliberation that occurs in response to the pre-issuance public comment in the context of regulations."'

On the other hand, rulings often are issued long after the enactment of the statute they interpret and, sometimes in anticipation of, during, or in the aftermath of the government's litigation with a taxpayer."' This practice, together with the government's direct interest in the outcome of tax litigation and the adversarial bias thereby engendered, weigh against deference-even Skidmore deference-for revenue rulings. Eliminating such deference would strengthen the courts' authority to check bureaucratic abuse arising from a desire to win.

The danger in taking this kind of bifurcated approach (Skidmore deference for regulations and no deference for rulings), not unlike the bifurcation effected under current law (Chevron deference for regulations and, apparently, Skidmore deference for rulings), is that it may simply result in the government's incorporating more interpretations in amendments to the regulations, rather than in rulings, in order to secure deferential review. The response is twofold. First, the suggested bifurcation will not create any more incentive for the government to act opportunistically than it has under the current law's bifurcation. Second, to the extent that the government begins to incorporate its interpretations opportunistically in regulations, this could lead to less, or even no, deference for such regulations under Skidmore. On balance, reverting to the traditional view that revenue rulings are not entitled to deference is preferable.333

The final question is whether the government should be able to argue in litigation that a taxpayer-friendly ruling or regulation should be denied effect. The issue can arise in either of two ways: (1) where the government argues that the ruling or regulation is invalid because it is inconsistent with the statute, or (2) where the government argues that, although the ruling or regulation is not invalid, it should nevertheless be rejected in favor of a newly proffered construction of the statute. In either case, as a matter of fairness, taxpayers who rely on such rulings or regulations should not face the possibility that the government will disavow them."

Efficiency concerns also favor the taxpayer. The uncertainty created by permitting the government to disavow its own rulings and regulations unquestionably impedes planning and makes transactions more difficult to consummate 335 Indeed, where the government disavows a ruling or regulation in court after the taxpayer consummates a transaction on the basis of it, the government is, in effect, making its new position operative on a retroactive basis-in violation of the spirit of the 1996 amendment to Code section 7805(b) 336

Thus, those decisions that have precluded the government from disavowing its interpretation of an ambiguous statute are a welcome development."' Legislation should be enacted confirming the validity of such corrective decisions and extending their reach so that the government is denied the authority to argue against an outstanding interpretation even when the government claims that the statute unambiguously renders the interpretation invalid.

As indicated, in Schleier,338 the Supreme Court concluded that courts are required to invalidate a revenue ruling if the ruling is found to be contrary to the plain meaning of a Code section at issue."9 Although the taxpayer in Schleier had not consummated the transaction on the basis of the ruling (it was issued after the taxpayer had completed the transaction),' the government nevertheless felt obliged to use its authority under Code section 7805(b)(8) to limit the Court's decision so that the declaration of the ruling's invalidity would have prospective effect only (i.e., would not become effective until the date of the Court's decision).341 However, the government's sensitivity to the plight of taxpayers who otherwise would have been adversely affected by the decision is not sufficiently reassuring to taxpayers in general. In other cases, the government may well prevail in litigation by arguing against one of its rulings and then refusing to limit the decision's retroactive effect.

In Schleier, the Court also raised, but did not answer, the question of whether a taxpayer-friendly regulation that is found to be invalid should similarly be disregarded.' However, Justice Scalia has indicated that such a regulation cannot constitutionally be permitted to have effect."' Under Justice Scalia's view, the Constitution vests the power to enact legislation in Congress, and therefore precludes an agency such as the Treasury from denying a statute the effect Congress intended.'"

Interestingly, the government's decision to limit the Schleier Court's invalidation of the revenue ruling by invoking its authority under Code section 7805(b)(8)"'I might raise constitutional questions as well. Once the Court declared that the ruling was contrary to the clear meaning of the statute, its conclusion became binding for all outstanding cases." In the absence of new legislation, it would be impermissible for the executive branch to alter or suspend the Court's decision.' Indeed, even with regard to new cases, any agency interpretation that is contrary to the clear meaning of the statute, as found by the Court, cannot be valid.348 Therefore, it would appear that Code section 7805(b)(8) may be unconstitutionally applied whenever, as in the aftermath of Schleier, the government seeks to revive a ruling declared invalid by the courts.' As a practical matter where the government revives a taxpayer-friendly ruling, any constitutional defect might well go unaddressed given the fact that no taxpayer is aggrieved." The lack of any challenge to the government's decision to make the Schleier holding prospective reflects this reality.351

As indicated, if the courts can resolve constitutional questions, fairness and efficiency would dictate that the government not be permitted to disavow its rulings or regulations. Once the government determines that adhering to its original position is no longer appropriate, the government should be required to effect any modification by revoking the ruling or regulation on a prospective basis. The government should not be permitted to leave a ruling or regulation intact and then ask the court to invalidate it. Thus, in policy terms, legislation denying retroactive effect to a court decision declaring a taxpayer-friendly ruling or regulation invalid would be salutary.352

The question then becomes whether-and if so, how-legislation accomplishing this objective could be fashioned on a constitutionally sound basis. One must consider two constitutional constraints. First, as indicated, the Supreme Court has concluded that the federal judiciary is constitutionally required to issue decisions on a retroactive basis?" Therefore, any decision determining the meaning of a statute becomes binding upon all pending and future cases, as well as the case under consideration." In adopting this approach, the Court has emphasized two rationales: the inappropriateness of discriminating among similarly situated litigants;"' and the institutional limitations inherent in the judicial function that require court decisions, unlike legislative decisions, to be forged in the concrete context of the pending case and all possible outstanding cases."" Second, the executive branch cannot constitutionally be given the authority to displace the judicially determined clear meaning of a statute."'

Given the first constraint, whether legislation could preclude the courts from giving retroactive effect to a decision overruling an agency interpretation is questionable. Nevertheless, the concern about discrimination does not seem insurmountable. After all, it is not invidious or unreasonable to distinguish between taxpayers who engage in a transaction before a ruling or regulation is declared invalid from those who do so after the courts decide the issue. Although the institutional-limitation rationale is somewhat more troubling, one can make a compelling argument that it should yield in order to avoid the harshness of applying a newly declared interpretation that is inconsistent with a ruling or regulation on which the taxpayer has relied. However, the second constraint presents a more formidable obstacle. That Congress could constitutionally mandate that an agency's misinterpretation be binding during the period prior to a court's decision declaring it invalid is doubtful (unless new legislation were enacted after the court's decision making the invalidated interpretation valid on a retroactive basis .. ). Thus, any legislation directing that agencies only give prospective effect to a court decision declaring a ruling or regulation invalid likely would not be sustained.

If, on the other hand, legislation were drafted to deny the Treasury funds to maintain any litigation designed to produce a result inconsistent with an outstanding ruling or regulation, the Treasury would not be placed in the position of rewriting the statute, but simply precluded from seeking a judicial interpretation contrary to its own outstanding interpretation. And if the legislation also denied the Treasury the authority to issue or revoke rulings or regulations on a retroactive basis,359 it would no longer be possi ble, as a practical matter, for a court to declare a taxpayer-friendly ruling or regulation invalid. Because this legislative proposal would neither authorize the Treasury to nullify the courts' interpretation of a statute nor interfere with the requirement inherent in the judiciary's function that courts make decisions on a retroactive basis, it would presumably survive constitutional analysis."

Nevertheless, other objections might be made to this proposal. First, it might be argued that the proposal seeks to achieve through indirect means what could not be achieved directly. In other words, because Congress cannot authorize an agency to rewrite a statute, an agency should not be permitted to accomplish that objective indirectly through the use of its appropriations power. Second, an objection might be made that the executive branch is charged with the responsibility of faithfully executing the laws and Congress cannot, having enacted a statute, deny the executive branch the funds to enforce it.361 Although not entirely free from doubt, the answer to these objections would appear to be that Congress ought to be able to qualify the rights and obligations that it creates.362 So, for example, if Congress were to determine that the Service should no longer audit income tax returns, it should be able to refuse to appropriate funds for this purpose without repealing the Code.363 And if Congress could create such a substantial impediment to the Code's enforcement, why should it be precluded from imposing the more limited constraint on enforcement of denying the government funds to seek a court decision that is contrary to an outstanding ruling or regulation?36 Thus, it would appear that there is no constitutional impediment to legislation that would deny the government funds necessary to maintain litigation seeking an outcome contrary to an outstanding ruling or regulation.

CONCLUSION

The transformation in deference has radically altered the role of the government in tax litigation. The government's declaration of victory by regulation in the Peterson/Simpson conflict, rather than following the traditional route of continuing to litigate, instantiates the government's enhanced authority. Given its bias as an adversary and the corresponding need for a disinterested check on bureaucratic abuse, a strong case can be made for reducing the level of deference the government enjoys in the tax context. The reduction should be made on an across-the-board basis. That is, regulations should no longer receive Chevron deference. Instead, courts should analyze their validity under the Skidmore framework. In the case of ambiguous regulations, a construction proffered by the government should receive no deference. Nor should the government enjoy any deference for a position it takes with regard to an ambiguous Code section during litigation. Lastly, deference should be denied revenue rulings as well. On the other hand, once the government has issued a taxpayer-friendly ruling (or regulation), fairness and efficiency require that it not be revoked retroactively. Thus, the government should not be permitted to pursue a position in litigation designed to produce an outcome contrary to one of its outstanding interpretations.

Mitchell M. Gans*

* Professor of Law, Hofstra University School of Law (a Fellow of The American College of Trust and Estate Counsel). I would like to acknowledge the assistance I received in writing this article. I want first to thank my dear friend, Jonathan Blattmachr, whose insights and encouragement were, as always, critical. My friend and colleague at Hofstra, Eric Friedman, always had time to listen and point me in a new, helpful direction. Bernard Jacob, another friend and colleague at Hofstra, took time to read the manuscript and make comments. I also want to acknowledge Linda Caller, another Hofstra colleague and friend, whose widely-cited articles on deference I cite at various junctures. Finally, I want to thank my research assistant Julie Tirella (class of 2001), who was helpful at an early point in the process, and my research assistant Steven Coran (class of 2002), for whose dedication to the project I remain very grateful. It all goes without saying that the errors are all mine.

Copyright American Bar Association, Real Property, Probate and Trust Law Section Winter 2002
Provided by ProQuest Information and Learning Company. All rights Reserved